Intelligent Fault-Tolerance Data Routing Scheme for IoT-Enabled WSNs

Wireless sensor networks (WSNs) have become one of the essential components of the Internet of Things (IoT). In any IoT application, different sensor-based devices gather data from physical objects and transmit the sensed information to the base station (BS). The BS analyzes this information depending on the sensor location. A high-performance intelligent WSNs is essential for any IoT-based application. In this article, high-performance intelligent WSNs are referred to as IoT-enabled WSNs. In IoT-enabled WSNs, fault occurrence probability is much more than traditional networks. Faulty nodes and broken links affect the reliability of the IoT-enabled WSNs. Various fault-tolerance algorithms enhance the network's reliability using multipath transmission, relay node placement, and backup node selection. However, these algorithms suffer from huge data transmission delays, packet overhead, and less detection accuracy. In this work, a multiobjective-deep reinforcement-learning (DRL)-based algorithm is proposed for fault tolerance in IoT-enabled WSNs. The main objective of this work is to detect the faulty nodes with high accuracy and less overhead. Furthermore, this work focuses on reliable data transmission after fault detection. Finally, a mobile sink (MS) is used for energy-efficient data gathering that significantly improves the network lifetime. Extensive simulations and theoretical analysis prove that the proposed algorithm outperformed as compared to the state-of-the-art algorithms in terms of fault detection accuracy (FDA), false alarm rate (FAR), false-positive rate (FPR), network lifetime, and throughput.